Classifications

• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03F—AMPLIFIERS
  • H03F3/00—Amplifiers with only discharge tubes or only semiconductor devices as amplifying elements
  • H03F3/20—Power amplifiers, e.g. Class B amplifiers, Class C amplifiers
• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03F—AMPLIFIERS
  • H03F1/00—Details of amplifiers with only discharge tubes, only semiconductor devices or only unspecified devices as amplifying elements
  • H03F1/32—Modifications of amplifiers to reduce non-linear distortion
  • H03F1/3223—Modifications of amplifiers to reduce non-linear distortion using feed-forward
  • H03F1/3229—Modifications of amplifiers to reduce non-linear distortion using feed-forward using a loop for error extraction and another loop for error subtraction
• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03F—AMPLIFIERS
  • H03F1/00—Details of amplifiers with only discharge tubes, only semiconductor devices or only unspecified devices as amplifying elements
  • H03F1/32—Modifications of amplifiers to reduce non-linear distortion
  • H03F1/3241—Modifications of amplifiers to reduce non-linear distortion using predistortion circuits
  • H03F1/3247—Modifications of amplifiers to reduce non-linear distortion using predistortion circuits using feedback acting on predistortion circuits

Definitions

• the present invention relates m general to radio frequency (RF) communication systems, and is particularly directed to an RF power amplifier linearization mechanism, that employs a spectral distortion measurement and differential combining scheme, which is operative to optimize the operation of a carrier cancellation combiner, and pre-distortion and feed-forward loops of the RF power amplifier, so that mtermodulation distortion produced at the output of the RF amplifier may be minimized.
• RF radio frequency
• IMDs mtermodulation distortion products
• a fundamental difficulty in linearizing an RF power amplifier is the fact that it is an inherently non-linear device, and generates unwanted mtermodulation distortion products (IMDs) .
• IMDs manifest themselves as spurious signals in the amplified RF output signal, separate and distinct from the RF input signal.
• a further manifestation of IMD is spectral regrowth or spreading of a compact spectrum into spectral regions that were not occupied by the RF input signal. This distortion causes the phase-amplitude of the amplified output signal to depart from the phase-amplitude of the input signal, and may be considered as an incidental (and undesired) amplifler-sourced modulation of the RF input signal.
• a straightforward way to implement a linear RF power amplifier is to build it as a large, high power device, but operate the amplifier at only a low power level (namely, at a small percentage of its rated output power), where the RF amplifier's transfer function is relatively linear.
• An obvious drawback to this approach is the overkill penalty -a costly and large sized RF device.
• Other prior art techniques which overcome this penalty include feedback correction techniques, feedforward correction, and pre-distortion correction. Feedforward and predistortion correction, however, are not limited m this regard.
• Feedback correction techniques include polar envelope correction (such as described m U.S. Patent No. 5,742,201), and Cartesian feedback, where the distortion component at the output of the RF amplifier is used to directly modulate the input signal to the amplifier in real time.
• Feedback techniques possess the advantage of self-convergence , as do negative feedback techniques in other fields of design.
• systems which employ negative feedback remain stable over a limited bandwidth, which prevents their application m wide-bandwidth environments, such as multi-carrier or W-CDMA.
• error present in the RF amplifier's output signal is extracted, amplified to the proper level, and then remjected with equal amplitude but opposite phase into the output path of the amplifier, so that (ideally) the RF amplifier's distortion is effectively canceled.
• predistortion correction a signal is modulated onto the RF input signal path upstream of the RF amplifier.
• the ideal predistortion signal has a characteristic that is the inverse or complement of the distortion expected at the output of the high power RF amplifier, so that when subjected to the distorting transfer function of the RF amplifier, it effectively cancels the distortion behavior.
• Either predistortion or feedforward may be made adaptive by extracting an error signal component in the output of the RF amplifier and then adjusting the control signal (s), in accordance with the extracted error behavior of the RF amplifier, so as to effectively continuously minimize distortion in the amplifier's output .
• One conventional mechanism for extracting the error signal component is to inject a pilot (tone) signal into the signal flow path through the amplifier and measure the amplifier's response.
• pilot tone A fundamental drawback to the use of a pilot tone is the need for dedicated pilot generation circuitry and the difficulty of placing the pilot tone within the signal bandwidth of the amplifier.
• pilot tone injection causes the generation of an unwanted spur; also, a piloted system is open-loop in the sense that the controller operates on the pilot and not the IMDs. Hence, the system only assumes that IMDs are being properly cancelled.
• the operation of a carrier cancellation combiner, and predistortion and feed- forward loops of an RF power amplifier are controlled to minimize IMD components at the output of the RF amplifier, by a spectral distortion measurement scheme that performs (Fast Fourier Transform (FFT) -based) spectral power measurements at a 'reference' signal port (associated with the RF input signal) , and a plurality of 'test' signal ports (associated with various parameter adjustment locations of the amplifier) .
• FFT Fast Fourier Transform
• Averaged FFT ' s of the data extracted from the reference port and data from the test ports by a digital signal processor based controller provide spectral information for each sampled signal set.
• the control data FFT is processed to establish a baseline, with which the test data FFTs are compared to generate adjustment signals for various control parameters, through which IMDs introduced by the amplifier are minimized.
• the performance of the RF amplifier is continuously monitored and the control parameters modified as necessary to compensate for drift in the amplifier's characteristics.
• the invention includes digitally controlled gain and phase adjustment circuits and a digitally controlled predistortion unit incorporated in the RF input signal path to a main RF amplifier.
• the predistortion unit may contain a work function-based vector modulator that is coupled to receive weighting coefficients from the controller. Since it contains any mtermodulation (spectral regrowth) distortion products (IMDs) introduced by the RF amplifier, the output of the amplifier is monitored as one of the test inputs to the controller .
• IMDs mtermodulation (spectral regrowth) distortion products
• the output of the mam RF amplifier is further coupled to a carrier cancellation combiner, a second input of which is coupled to a feed forward path from the RF input signal port.
• the feed forward path from the RF input signal port includes a fixed delay and a variable delay unit, that serve to substantially equalize the propagation delay of the signal path through the mam RF amplifier, and thereby provide proper phase alignment of the signals applied to the carrier cancellation combiner.
• the output of the carrier cancellation combiner is coupled to gam and phase adjustment circuits of a feed forward error amplifier, the output of which is rem ected into the output path of the RF amplifier, for feed-forward IMD cancellation.
• the DSP-based controller performs various spectral distortion measurement operations, and error minimization algorithms such as, but not limited to power or least mean squared minimization, to control variable ga and phase shift components both the mam RF amplifier and error amplifier signal paths, for the purpose of optimally canceling IMD components at the RF output port. It also generates work function-based predistortion control signals derived from respectively different work functions of the instantaneous amplitude of the RF input signal, so as to predistort phase and amplitude components of the RF input signal to the ma RF amplifier.
• the DSP controller is operative to process the spectral data samples derived from the control and test ports, and then adjusts the parameters of the control components as necessary to compensate for ma RF amplifier distortion.
• the controller FFT
• FFT-processmg the data allows spectral energy m one or more frequency bins, other than the portion of the spectrum in which the carrier components are located, to be selectively discarded, to avoid erroneously influencing the carrier cancellation loop.
• Optimal (maximum) cancellation of carrier components is accomplished by adjusting the gam and phase adjustment circuits m the RF input signal path and the adjustable delay unit, so as to maximize the expression:
• N the total number of carriers
• Ci is the average power of carrier Ci of N.
• Ri is the average power of residual carrier Ri of N.
• N i ⁇
• Di is the distortion component i of N detected spectral components resulting from the subtraction operation.
• the FFT of the control data is subtracted from the FFT of the of RF output signal monitored downstream of the feed forward reinjec ion port.
• the gain and phase adjustment circuits for the feed forward error amplifier are then adjusted to minimize the following expression:
• Di is the distortion component i of N detected spectral components resulting from the subtraction operation.
• Figure 1 diagrammatically illustrates an RF power amplifier spectral measurement and distortion correction scheme in accordance with an embodiment of the invention.
• Figures 2-5 are respective spectral diagrams associated with the operation of the RF power amplifier arrangement of Figure 1.
• the new and improved RF power amplifier spectral measurement and distortion correction mechanism resides primarily m a prescribed arrangement of conventional RF circuits, associated digital signal processing components and attendant supervisory control circuitry, that controls the operation of such circuits and components.
• the configuration of such circuit components, and the manner in which they interface with other communication system equipment have, for the most part, been illustrated m the drawings by a readily understandable block diagram, which shows only those details that are pertinent to the present invention, so as not to obscure the disclosure with details which will be readily apparent to those skilled in the art having the benefit of the description herein.
• the block diagram illustration is primarily intended to show the components m a convenient functional grouping, whereby the invention may be more readily understood.
• an RF power amplifier spectral measurement and distortion correction scheme m accordance with the present invention is diagrammatically illustrated as comprising an RF input port 11, to which an RF signal RF ⁇ n to be amplified is coupled.
• RF input port 11 is coupled over an input signal path 13 to a mam RF power amplifier (or basic power module (BPM) 20, whose non-linear spectral distortion (IMD) introducing behavior is to be compensated.
• the input signal path 13 to RF amplifier 20 includes digitally controlled gam and phase adjustment circuits 14 and 15, respectively, and a digitally controlled predistortion unit 16.
• the digitally controlled predistortion unit 16 may contain a work function-based vector modulator, that is coupled to receive a set of weighting coefficients w 0 , w 1# w 2 , ..., w N , supplied by a performance monitoring and parameter- updating digital signal processor (DSP) -based controller 50.
• DSP digital signal processor
• the DSP controller 50 executes spectral distortion measurement and error minimization algorithms (to be described) for adjusting the pre-distortion properties of predistortion unit 16, and also controls the digitally controlled gam and phase adjustment circuits 14 and 15 of the input signal path, digitally controlled ga and phase adjustment circuits 44 and 45 of a feed forward loop 43 to an error amplifier 40, and an adjustable delay unit 22 that is coupled m a feed forward path 23 to a carrier cancellation combiner 60, as will be described.
• control lines 51 from controller 50 are denoted by subscripts associated with the respective components being controlled.
• the output of the RF power amplifier 20 is coupled through a downstream delay unit 17 and circulator 18 to an RF output port RF out .
• An RF output directional coupler 19 is coupled to the output path of the amplifier by means of a feed forward reinjection directional coupler 47.
• RF output directional coupler 19 serves as a test signal port that is employed to extract a signal 'OUT', representative of the composite amplified RF signal, including intermodulation (spectral regrowth) distortion products (IMDs) introduced by the RF amplifier, and reductions therein due to the feed forward injection.
• This extracted RF output signal (OUT) is coupled from the directional coupler 19 to a first input 31 of a DSP-controlled switch 30.
• the RF input port 11 is coupled through a directional coupler 21 to a second input 32 of controlled switch 30.
• the output 35 of the switch 30 is coupled to a first input 71 of a mixer 70, a second input 72 of which is coupled to receive an IF frequency provided by a local oscillator 75.
• the mixer 70 is operative to down-convert the output of the switch 30 to baseband.
• This baseband signal is then filtered in a bandpass filter 76, digitized by a high speed analog-to-digital converter (ADC) 77, and then stored in a buffer memory 79 for analysis by the DSP controller 50, as will be described.
• ADC analog-to-digital converter
• the steering operation of switch 30 and read/write control of memory 79 are controlled by a control link 52 from the DSP 50 controller.
• the feed forward path 23 to the carrier cancellation combiner 60 (which may be configured as a Wilkinson combiner) includes a fixed delay line 24 coupled in series with the variable delay unit 22 from the RF input path 13 to a first input 61 of the carrier cancellation combiner 60.
• the second input port 62 of the carrier cancellation combiner 60 is coupled via a directional coupler 25 to the output of the RF amplifier 20.
• the delay is employed to substantially equalize the propagation delay of the feed forward path with the RF signal path through the RF amplifier to a second input port 62 of the RF carrier cancellation combiner 60 and ensure phase alignment of the signals applied to the combiner.
• the carrier cancellation combiner 60 cancels RF carrier components at its output, so as to produce a signal 'Eamp' representative of the IMD portion of the output of the RF amplifier 20.
• the signal Eamp is coupled to respective digitally controlled gain and phase adjustment circuits 44 and 45 to the error amplifier 40, whose output is reinjected into the output path of RF amplifier 20 by the directional coupler 47 installed upstream of directional coupler 19, as described above.
• the signal Eamp is extracted via a further test port directional coupler 65, installed with the feed forward loop 43 at the output of the carrier cancellation combiner 60, and coupled to a third input 33 of the controlled switch 30.
• An additional directional test port coupler 46 is coupled to the path from the directional coupler 25 at the output of the mam RF amplifier 20, so as to provide a further signal 'BPM' representative of the amplified output of amplifier 20 to fourth input 34 of controlled switch 30.
• the DSP-based controller 50 uses various spectral distortion measurement operations, and error minimization algorithms (e.g., power or least mean squared minimization) for controlling variable gam and phase shift components m both the mam RF amplifier and error amplifier signal paths, for the purpose of optimally canceling IMD components at the RF output port RF ouC . It also generates work function-based predistortion control signals derived from respectively different work functions of the instantaneous amplitude of the RF input signal to predistort phase and amplitude components of the RF input signal to the ma RF amplifier .
• error minimization algorithms e.g., power or least mean squared minimization
• the DSP controller 50 executes what are effectively (Fast Fourier Transform (FFT) -based) spectral power measurements in portions or bins of the frequency spectrum of each of the monitored reference port signal 'SRC, and test port signals 'OUT', 'Eamp', and 'BPM', downconverted to baseband, bandpass filtered, sampled and then stored as a picture or 'snapshot' of the spectral composition of the entire band.
• FFT Fast Fourier Transform
• Averaged FFT ' s of reference data samples (associated with the SRC port) and respective sets of test data samples (associated with the OUT, Eamp and BPM ports) are performed to obtain the spectral information for each sampled signal set of a respective spectral snapshot.
• the control (SRC) data is processed to establish a baseline, with which the test data is compared to produce adjustment signals for the various control parameters, and thereby minimize IMDs introduced by the main RF amplifier. With the control parameters initially corrected, the performance of the main RF amplifier is thereafter continuously monitored and the adjustable parameters as necessary to compensate for any drift in the amplifier's characteristics.
• FFT Fast Fourier transform
• FFT processing of the carrier cancellation test data will typically produce a spectral distribution as shown m the spectral diagram of Figure 3 , which includes (reduced amplitude) residual carrier components 301 and 302 (respectively associated with the carriers 201 and 202 of Figure 2), as well as spurious (IMD) components shown m spectral regions or bins 303 and 304 outside the spectral region containing the carriers, the energy content of which should not contribute to carrier cancellation.
• N the total number of carriers
• Ci is the average power of carrier Ci of N.
• Ri is the average power of residual carrier Ri of N.
• maximizing the expression (1) has the effect of reducing the average energy the identified residual carrier components at the output of carrier cancellation combiner 60 to as close to zero as possible (optimal carrier cancellation) , without influencing the maximization operation to energy in spurious components m frequency bins other those containing the carriers.
• this provides a direct measure of carrier cancellation (with both IMDs and carriers being measured simultaneously) , the system effectively 'knows' whether or not it is functioning properly - which is not achievable with conventional correlator-based or power-detector energy minimization approaches .
• Such conventional power detector energy minimization and correlator based schemes are subject to input drive power and cannot resolve whether power is present or not.
• the present invention obviates the need for additional circuitry to 'interrupt' the control loop under this condition, and eliminates zero- input drift.
• the use of the adjustable delay element 22 allows the invention to detect and optimally tune the delay for any signal misalignment, thereby maximizing carrier cancellation.
• FFT processing of the pre-distortion test data will typically produce a spectral distribution as shown in the spectral diagram of Figure 4, having amplified carrier components 401 and 402 (respectively associated with the carriers 201 and 202 of Figure 2) , as well as IMD components shown in regions 403 and 404, that lie outside the spectral region of the carriers.
• N i ⁇ where Di is the distortion component 1 of N detected spectral components resulting from FFT BPM - FFT SRC .
• EAMP may be employed as the test port m place of the BPM port, since each contains the same IMD information.
• the present invention enables IMD ' s to be minimized m the presence of carrier, it avoids the problems of carrier energy removal -based approaches, which require some amount of carrier energy removal m order to detect changes m the pre-distortion circuitry, and therefore suffer performance monitoring degradation as more carrier energy is leaked.
• the DSP controller 50 processes the SRC data to identify any carriers present m the RF input signal ( Figure 2) and the noise floor.
• FFT processing of the feed-forward data (OUT) will typically produce a spectral distribution as shown in the spectral diagram of Figure 5, including amplified carrier components 501 and 502 (respectively associated with the carriers 201 and 202 of Figure 2) , as well as (reduced amplitude) IMD components shown m regions 503 and 504, that lie outside the spectral region of the carriers.
• Di is the distortion component l of N detected spectral components resulting from FFT 0UT - FFT SRC .
• the spectral reduction mechanism of the invention is particularly advantageous with respect to pilot tone based systems that measure amplitude and phase differences of output-delay line and feed-forward path.
• the invention inherently accounts for changes in the amplifier that cause 'targets' of the control loop to drift (and mandate periodic re-calibration of pilot tone systems), and is therefore effectively self -calibrating .
• the invention is wide bandwidth, whereas pilot tone approaches operate only a portion of the band, or operate out-of-band and simply assume that proper cancellation will be performed in the actual band of interest .
• pilot tone based receiver and pilot tone based energy reduction schemes are unable to directly measure IMD performance or detect spurs, and therefore may leave a spur that is ' out-of -spec ' to prevent the amplifier's true performance from being seen.
• the present invention is able to minimize IMD ' s to within a given specification and can directly measure IMD performance .
• the FFT subtraction operations may be replaced by applying the test data to the signal carrier used by the control port.
• the summation operation of each of equations (2) and (3) would then ignore signals already identified at the control port .

Landscapes

• Engineering & Computer Science (AREA)
• Power Engineering (AREA)
• Physics & Mathematics (AREA)
• Nonlinear Science (AREA)
• Amplifiers (AREA)

Applications Claiming Priority (3)

Publications (3)

Family

ID=24036819

Family Applications (1)

Country Status (8)

Cited By (3)

Families Citing this family (34)

Citations (1)

Family Cites Families (14)

• 2000
  • 2000-02-25 US US09/511,877 patent/US6275106B1/en not_active Expired - Lifetime
• 2001
  • 2001-02-14 DE DE60139547T patent/DE60139547D1/de not_active Expired - Lifetime
  • 2001-02-14 AT AT01912738T patent/ATE439700T1/de not_active IP Right Cessation
  • 2001-02-14 JP JP2001562824A patent/JP4913304B2/ja not_active Expired - Fee Related
  • 2001-02-14 AU AU2001241485A patent/AU2001241485A1/en not_active Abandoned
  • 2001-02-14 KR KR1020027010925A patent/KR20030010583A/ko not_active Application Discontinuation
  • 2001-02-14 WO PCT/US2001/004737 patent/WO2001063748A1/fr active Application Filing
  • 2001-02-14 EP EP01912738A patent/EP1262017B1/fr not_active Expired - Lifetime

Patent Citations (1)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events